Diet Plan or Exercise Plan: Which One Would You Still Be Doing in a Year?

Head-to-head 1: Racette et al. 2006. Racette SB, Weiss EP, Villareal DT, Arif H, Steger-May K, Schechtman KB, Fontana L, Klein S, Holloszy JO; Washington University School of Medicine CALERIE Group. One year of caloric restriction in humans: feasibility and effects on body composition and abdominal adipose tissue. J Gerontol A Biol Sci Med Sci 2006;61(9):943–950. doi:10.1093/gerona/61.9.943. PMID 16960025. Randomized controlled trial, 1 year: 48 healthy, nonobese men and postmenopausal women aged 50–60 (BMI 23.5–29.9; mean 27.3 ± 0.3; exercising no more than twice a week at entry), randomized 2:2:1 to calorie restriction (n = 19), exercise designed to produce a similar energy deficit without eating less (n = 19) or a healthy-lifestyle control (n = 10); 46 (96%) completed. Both prescriptions ramped from about 16% to 20%: 524 ± 33 kcal/d less food versus 563 ± 18 kcal/d of extra exercise; matching the 20% by exercise 'required approximately 90 minutes every day'. Weekly dietitian and psychologist meetings for 6 months (diet); trainers working individually with exercisers; weekly weigh-ins for both. Measures: diet adherence as %CR from doubly labeled water with DXA-derived changes in body energy stores; exercise adherence from heart-rate monitors plus self-reported unrecorded sessions. Results: 11.5 ± 2.1% CR averaged over the year, lower in the second half; exercisers averaged 5.8 ± 0.6 sessions/wk of 62.5 ± 4.3 min and 317 ± 39 kcal/d, 58.7 ± 6.7% of prescription (abstract: 59%), likely an underestimate because heart-rate data were missing for 12.8 ± 4.0% of sessions; the exercise figures are year averages, and the paper gives no time course for exercise adherence. Weekly weigh-in attendance 84.6 ± 2.8% versus 85.0 ± 3.2% of intervention weeks. One dropout per active group (diet: inability to adhere to the prescription, month 6; exercise: unrelated medical reasons, month 9). Adverse events: hunger and constipation (diet); muscle soreness and joint pain (exercise). 7-day food diaries were 'consistently lower than DLW-determined intake for all groups'; in the diet group, diaries read 1,739 kcal/d at 3 months and 1,737 at 1 year while DLW-derived intake rose from 2,156 (months 3–6) to 2,259 (6–9) and 2,375 (9–12), a comparison of Table 3 rows made by FitChef, not by the authors. Graduate-school education: 58% (diet) and 47% (exercise). The authors note that the people who enrolled 'were ready to make lifestyle changes' (a few control participants lost weight with minimal contact). Funding: NIH Cooperative Agreement 5-U01-AG20487 and grants RR00036, DK20579, DK56341, AG00078; no industry funding or authorship stated.

Head-to-head 2: Foster-Schubert et al. 2012 (NEW trial). Foster-Schubert KE, Alfano CM, Duggan CR, Xiao L, Campbell KL, Kong A, Bain C, Wang CY, Blackburn G, McTiernan A. Effect of diet and exercise, alone or combined, on weight and body composition in overweight-to-obese post-menopausal women. Obesity (Silver Spring) 2012;20(8):1628–1638 (online 14 April 2011). doi:10.1038/oby.2011.76. PMID 21494229. 12-month, 4-arm randomized controlled trial (Seattle, 2005–2009): 439 sedentary postmenopausal women aged 50–75 with BMI ≥25 (≥23 for Asian-American women) exercising under 100 min/wk; diabetes and other listed conditions excluded; about 85% non-Hispanic White and 65% college graduates. Arms: diet (N = 118), exercise (N = 117), diet + exercise (N = 117), control (N = 87); 399 (91%) completed. Diet: 1,200–2,000 kcal/d, under 30% of energy from fat, 10% weight-loss goal; individual goal-setting, then weekly groups of 5–10 for 6 months, then twice-monthly contact. Exercise: 45 min or more of moderate-to-vigorous exercise 5 d/wk (225 min/wk), at least three supervised facility sessions a week and the rest at home; logs reviewed weekly; staff contacted women missing targets. Adherence measures: diet by session attendance, food-journal completion and self-weighing (and weight loss); exercise by daily logs, with home sessions self-reported. Contact time with study staff 'varied to some degree by intervention group' (least, by design, for the controls), which the authors name as a possible source of variability in weight loss. Results: both diet arms attended an average of 27 diet sessions (86%, the Results figure). Exercise-only women achieved 80% of the 225-min/wk target. Dropouts: diet 13, exercise 11, diet + exercise 9, control 7 (counts; no between-arm test); main reasons: dissatisfaction with randomization, work or family demands, or injury or other medical issues unrelated to the intervention. Weight: diet alone −7.2 kg (−8.5%), exercise alone −2.0 kg (−2.4%); setting the smaller loss beside the dropout counts is FitChef's reading. The FFQ found no significant change in total calories in the diet arms versus controls ('not as robust for estimating this parameter'); the authors cite the weight loss as evidence the dieters ate less. Funding: National Cancer Institute (R01 CA102504, U54-CA116847; KFS NIH 5KL2RR025015-03; AK R25 CA94880); disclosure: 'Nothing to disclose'.

Different yardsticks. In both trials each plan was scored with its own tool, doubly labeled water against heart-rate monitors in the first and session attendance against logged minutes in the second, so each pair of figures is similar rather than a like-for-like test, and neither trial supports subtracting one figure from the other or naming a winner. Weekly weigh-in attendance in the first trial (84.6% versus 85.0%) was recorded the same way for both plans, as were the dropout counts in both trials. The diet group's share of target in the first trial is FitChef's derivation (11.5% against the final 20% = 57.5%; against the time-weighted 19% = 60.5%); the paper reports no percent-of-prescription for the diet group, and the 58.7% belongs to the exercise group only.

The pooled scoreboard: Lemstra et al. 2016. Lemstra M, Bird Y, Nwankwo C, Rogers M, Moraros J. Weight loss intervention adherence and factors promoting adherence: a meta-analysis. Patient Prefer Adherence 2016;10:1547–1559. doi:10.2147/PPA.S103649. PMID 27574404. 27 studies published January 2004 to August 2015 (20 RCTs and 7 observational intervention studies; 6,803 participants). Adherence 'was defined as completion of the weight loss program or, in certain cases, was assessed by the level of consistency with the weight loss intervention of interest.' Diet-only 63.73% (N = 1,933) versus physical-activity-only 50.21% (N = 1,388): RR 1.27 (95% CI 1.19–1.35). Table 4's subgroup Ns sum past the 6,803 participants, so they count adherence measures, not people, and the comparison is between different programs, not arms of the same trials. The authors' possible explanations, untested: 'poor recall or false information' where diet adherence was self-reported, and more weight loss from restrictive diets than from exercise. Their stated limitation: pooling heterogeneous studies 'does introduce a level of uncertainty ... due to possible sampling errors or unmeasured covariates.' Context only (support and supervision are a separate question from plan type): supervised programs 68.6% versus self-monitoring programs 41.5%; programs under 12 months 69.9% versus 12 months or longer 53.0%. One row of the review's Table 3 (van Gool et al. 2006; 18-month RCT, 238 participants, overweight and obese older adults) reported attendance of 52.3% in the diet arm (n = 137) and 49.9% in the physical-activity arm (n = 133); it is cited here only as a row of the review's table. The authors report no conflicts of interest.

Diets lose people too: Dansinger et al. 2005. Dansinger ML, Gleason JA, Griffith JL, Selker HP, Schaefer EJ. Comparison of the Atkins, Ornish, Weight Watchers, and Zone diets for weight loss and heart disease risk reduction: a randomized trial. JAMA 2005;293(1):43–53. doi:10.1001/jama.293.1.43. One-year, single-center randomized trial of four popular diets in 160 adults aged 22–72 with overweight or obesity (BMI 27–42) and known hypertension, dyslipidemia or fasting hyperglycemia; no exercise arm. 34 (21%), 61 (38%) and 67 (42%) had not completed the study at months 2, 6 and 12; discontinuation trended higher on the more extreme diets (Atkins 48%, Ornish 50%) than on the moderate ones (Zone and Weight Watchers, 35% each; P = .08, not significant). The most common reasons for leaving were that the assigned diet was too hard to follow or not yielding enough weight loss.

Keeping weight off is a separate question: Varkevisser et al. 2019. Varkevisser RDM, van Stralen MM, Kroeze W, Ket JCF, Steenhuis IHM. Determinants of weight loss maintenance: a systematic review. Obesity Reviews 2019. doi:10.1111/obr.12772. Strong evidence that an increase in physical activity predicts weight loss maintenance, and moderate evidence that baseline activity does not. A maintenance finding, cited in the FAQ; it does not show that exercise is easier to stick to.

What this page does not claim. Two trials, one of them small; adults aged 50–75 (healthy, nonobese men and women in one; postmenopausal women with overweight or obesity in the other); both plans came with heavy professional support in both trials, and the second trial's authors say applicability to less intensive community programs 'is not as clear'; one year only. The did-not-quit-more observation and the diary-versus-measurement comparison are FitChef's readings of the published tables, not analyses the authors ran. Nothing here concerns disease risk, and none of it is advice on which plan to choose.